Neither auto companies nor tech companies can come up with winning mobility offerings on their own. Instead, they will have to work together to create products, services, and business models to meet the needs of users.
For more than a century, the business of moving people has been dominated by the automobile, with major automakers manufacturing vehicles and selling them to consumers. It has always been a capital-intensive business with extremely low margins. Now this business is on the cusp of dramatic technological change. A new vision has emerged where mobility is a service, enabled by automated vehicles (AVs) that are accessed on demand and operate without the need for human drivers. The new entrants — which include tech giants Google and Apple, ride-hailing pioneers Uber and Lyft, and suppliers such as Nvidia and Mobileye — have technological expertise, ready access to engineering talent, and deep venture-capital pockets on their side. Many observers, including prominent Wall Street analysts, think tech challengers from Silicon Valley will have the upper hand over auto companies.
In our view, however, that outcome is by no means assured. The most important question isn’t “Who will win?” but “How do you win?” Having studied the history of supply chains, the dynamics of competition and value migration, and the diffusion of new technologies in the automotive industry (and other manufacturing sectors), we think the future of mobility isn’t autos versus tech, but autos plus tech, based on collaborations that weave together products, services, and business models to meet the needs of individual users across wide-ranging use cases.
Neither auto companies nor tech companies have the ability to come up with winning mobility offerings on their own. Auto companies will never be that good at running service-based businesses or monetizing data; technology companies are unlikely to enter the complex and low-margin business of vehicle manufacturing. And not every collaboration (which may include joint ventures, acquisitions, and policy coalitions) will prosper. Even if it’s many years before AVs are driving on our streets, the planning needs to happen now. The various players will need to sort out a variety of complex issues, including how to handle concerns about safety, who controls the data, and who reaps the value. In this article, we will look at some of the possible scenarios.
The Power of Systems Integrators
In recent decades, the dominant narrative of technological disruption has centered on how new entrants introduce early products into rapidly developing markets, stay off the radar screens of incumbents, and then become industry-dominating leaders. This, of course, was what happened in personal computers, when Microsoft and Intel attacked IBM. However, value migration is neither inevitable nor random but a direct consequence of the decisions by incumbents and new entrants.1 In the auto industry, automakers (not disrupters) have managed to preserve their dominant share of value for over a century by being systems integrators. Knowing how to bring the various technologies and functions together — everything as wide-ranging as product design, supply chain management, manufacturing, and after-sales support — has been the crucial element in maintaining this status. Some analysts doubt that the auto companies (known as original equipment manufacturers, or OEMs) will be able to maintain this position in the years ahead. However, we think systems integration capabilities will continue to give auto companies an important edge.
In recent decades, auto companies have shown that they can buy most of their components and still maintain the lion’s share of the value in their supply chains.2 They protect this value in part by validating the safety of vehicles and assuming legal liability for products that can cause death, injury, and property damage.3 Any predictions that tech companies have an opportunity to overtake the auto companies in the era of automated vehicles must be able to explain how the new players will be able to take control of the supply chain and its associated responsibilities.
When people talk about automated vehicles, a variety of different functions get lumped together. At the macro level, AVs aim to navigate the environment without the need for human drivers, promising to make driving safer, more efficient, and less tiring with the help of sensors (including cameras; radar; and light, detection, and ranging technology, known as lidar), machine learning algorithms, and high-resolution maps. However, the large-scale introduction of vehicles that can self-drive without human intervention in any environment is probably decades away, and it will call for the best efforts of both auto and tech companies.
A critical enabler of this AV future will be the creation of a new and more advanced supply chain — or data value chain4 — that can draw on the expertise of both new companies with digital expertise and incumbents. The goal will be to coordinate the suppliers providing the electromechanical features with those producing the digital elements (including software, sensors, data processing, and navigation) to make autonomous cars a reality. In this task, incumbent OEMs with long-standing systems integration capabilities will have a huge advantage.
What’s needed isn’t simply the ability to specify, purchase, and coordinate the complex logistics for diverse technological components that OEMs have already mastered. Those seeking to manage supply chains must also have substantive knowledge of the core technologies so that they can meaningfully engage with suppliers. For some, this will mean hiring engineers, acquiring startups, partnering with tier 1 suppliers, and even participating in the data value chain as customers. For example, Ford recently acquired a company called Quantum Signal, a small robotics company that had worked with the military to develop software for remotely controlling vehicles and simulating vehicle testing.5
Will Tech Companies Manufacture Their Own Vehicles?
Given that the ability to integrate hardware and software to produce safe and comfortable products will be key, what will prevent tech companies such as Apple or Google’s Waymo from entering the market as systems integrators and even designing and manufacturing their own vehicles? In our view, tech companies don’t currently have the requisite manufacturing capability — and given the low margins, it’s not likely anyone would spend money to develop or acquire this capability. Automobiles are complicated products that need to draw on an array of technologies and then perform on a variety of dimensions. If nothing else, Tesla’s recent experience getting its Model 3 production up to speed highlights how challenging it is to produce high-quality vehicles at scale. Contract manufacturing doesn’t appear to be a realistic possibility; Magna International, a major automotive supplier based in Ontario, is currently the only company with significant contract manufacturing capabilities.
Rather than designing and manufacturing their own vehicles, tech companies are more likely to try to provide the operating system for many different vehicle brands and rely on others to do the vehicle assembly. (This supports our viewpoint that the future will be autos plus tech.) Currently, Waymo’s technologies (including a proprietary lidar design that it decided was worth fighting with Uber in court to protect) are being refined in fleets of test vehicles built by Fiat Chrysler and Jaguar Land Rover and are expected to appear in other vehicles soon. Other collaborations on AVs are also underway (including Ford and VW investing in Argo AI, and General Motors and Honda in Cruise) as automakers strive to hold on to their role as systems integrator.
Getting Auto and Tech Companies to Work Together
Auto companies know vehicles, and tech companies know software. Together they are well equipped to produce the AVs of the future. But serious attempts at collaboration are certain to bring up thorny issues. Chief among them: Who owns the data? AVs generate huge amounts of potentially valuable vehicle and consumer data. OEMs, for their part, have been scrambling to learn how to combine their traditional electromechanical knowledge of vehicle components with digital and electronic expertise. Tech companies already know how to manage and monetize data. While auto OEMs could try to monetize customer data on their own with new services, and tech companies could attempt to do things like manage the operations end of robo-taxi fleets, it would be easier if the two groups worked together and shared the value generated by the data.
Another huge issue will be working with regulators, insurance companies, and municipalities to make data available for public purposes. Auto companies have an advantage here, having had decades of experience working to meet societal expectations. In fact, we believe that the tech industry’s ethos of “move fast and break things” will be a liability with consumers and governments. The ongoing controversy in Los Angeles over how much data the operators of bike- and scooter-sharing systems need to share with the city offers a preview of the challenges.6
Exactly which AV business model will prosper in the coming years is an open question. Being a first mover like Waymo may have advantages: The sooner a credible player is able to deploy functional AVs, the more data it will collect, permitting it to improve its automated driving algorithms and deploy more vehicles. However, we think the pace of AV adoption will be slow, allowing fast followers like Cruise and Argo AI to remain competitive. Under any scenario, regulation and the need for a “social license” will play a key role in determining how and when new self-driving capabilities are deployed. We see this favoring “autos plus tech” as the strategy most likely to succeed in moving automated vehicles into the mainstream.